Compute forward kinematics for a batch of UR5e joint configurations using PyRoKi.
Parameters
batch_joint_states
np.ndarray | List[List[float]]
required
Batch of joint configurations of shape (N, 7). Can also be a path to a .npy file.
Returns
dict with keys:
output — Array of end-effector poses (N, 7) in wxyz_xyz format (quaternion wxyz + position xyz)latency_ms — Processing time in milliseconds
Example
from grid_cortex_client import CortexClient
import numpy as np
client = CortexClient()
batch_joint_states = np.random.uniform(-1.0, 1.0, size=(20, 7)).astype(np.float32)
result = client.run(
model_id="ur5e-pyroki-fk",
batch_joint_states=batch_joint_states,
)
poses = result["output"]
print(poses.shape) # (20, 7) — wxyz quaternion + xyz position
